One of the home improvement shows featured a device that was meant to install under a bathroom sink. It was a pump connected to both hot and cold water lines, and would pump water from the hot line to the cold line. The home owner would run the pump for a short while before using the water. The intent was to bring hot water up from the water heater (located on the other end of the house) so the homeowner would have hot water almost immediately (including that bathroom’s shower).
It seems to me that if this device could be set to automatically cycle on and off, it would also be useful to prevent frozen pipes - in that one circuit.
Does moving water actually freeze at a lower temperature? How much lower? Like, imagine you had a shaded circular pipe free-standing in the air with a pump pushing water around it. Without the pump running, the water will freeze if the air temp drops below freezing. If you run the pump (and let’s assume for the moment that the pump adds no heat to the system; it would obviously add some), at what temperature will the water freeze? My contention is that it’s the same temperature. There’s no additional heat being added to the system, and there’s nothing magic about water moving that will keep it from freezing. The reason moving water in a pipe or a pond doesn’t freeze as quickly is that warmer water is moving in to keep things above freezing.
The reason that the bubbles moving water near the dam keeps it from freezing is that the portion of the dam pond that’s below the frost line is at non-freezing temperatures. It’s exactly the same reason that running the faucets keeps things from freezing. Water in the ground below the frost line is too warm to freeze.
Perhaps not =) But in the blizzard in December in um … 1983? 84? I was living on the shore of Lake Ontario in Hilton NY and we were snowed in for 10 days and the lake was frozen solid out about half a mile, it was not smooth but rough because of the waves spraying water onto the edges of the freeze and causing sort of chunks and areas of almost frozen froth. Was really cool =) And I can remember in 74 being up in Picton Ontario and my dad driving out to Grapey Island in our 74 Chrysler NYer on the ice.
I always conceive of it as being the reverse of running cold water onto a burn to cool it down- running above freezing water through a pipe is going to keep it warmer than it otherwise would be if it was still, and thereby prevents freezing. Basically what @UltraVires is saying- it’s “heating” it in a relative sense- if you’ve got 45-50 degree water slowly flowing through a pipe, you’re adding heat into the pipes and surrounding area, thereby (hopefully) keeping the whole thing above freezing.
In our house, the water lines from the main are all underground and therefore protected more or less from freezing, but once they get into the house, they mostly go through the crawl space underneath, which isn’t heated, so they could potentially freeze. So running the pipes keeps those pipes from freezing. (that said, I’ve found that sealing the crawlspace vents with cardboard and tape does a fantastic job of keeping the temperature under there high enough when there’s a big cold snap. We used to get some outer wall pipes that would freeze when the temp got below about 10-12 F, but when I seal the crawlspace off, they’re good to at least -2, as I found out last February (I live in North Texas).
If an ice plug develops, maybe the pressure differential between the street side and the faucet side of the plug allows a bit of water to flow around any ice plug that develops. This might prevent the plug from sticking to the walls of the pipe. Or maybe just the pressure differential prevents the ice plug from being able to fully form in the first place. The pressure would keep pushing away the ice crystals that are trying to close up the hole.
Ice, being a solid, has some amount of shear modulus, while water, being a liquid, has very little. That is, ice is effectively thickening the walls of the pipe, so that the water pressure has to crack the ice as well as the pipe. Where there’s no ice, it’s the pipe has to contain the pressure all by itself.
Leaving faucets barely open is effectively creating a fuse. It creates a weak point in the pressure vessel created by an ice blockage. As the pressure increases, the water flow out the tap will as well. That means when your pipes are in the process of freezing, do not keep adjusting the faucet to minimize the drips. You’d be clamping the pressure relief just as the pipes need it the most.
I don’t know why people think there is only one mechanism in play…
Clearly, flowing water is going to warm the pipes. That will slow the rate of freezing down.
Leaving a tap open will also provide a pressure relief, which may prevent the pipes from bursting if they do freeze.
Also, the flowing water may prevent the pipes from freezing solid, if they do start to freeze.
The amount seems negligible, though. A drop of water is about 0.05 ml. Say a steady drip is twice a second, for 0.1 ml/s. If the incoming water is 10 C, then at most it can give 10 K to the pipes or any forming ice. That’s 0.1 ml/s * 1 g/ml * 4.2 J/g-K * 10 K = 4.2 W. That’s not a whole lot of warming power at play.
Also, as I said, because water at 4C is at its densest - so ice being less dense, floats and forms a crust that tends to insulate the body of water.
I have seen water flows at -40C. Where the water goes over a falls or rapids, it will stay open. slower flowing water allows the ice to form where the motion is less, and it builds out from there (see example above that it forms along the shore of Lake Ontario) Once water stops moving long enough to lose the heat it has (80cal/g) to freeze solid, it will - so the spray from a waterfall will freeze against the scenery around it. It may eventually form a bridge and cover to hide the flowing water, which is 4C or so.
And yes, the deeper ground does not lose too much heat, if it’s covered by a deep layer of water at 4C. Whatever heat it absorbed in the summer will leak out over the winter into the water.
For water flowing in pipes - it is coming from somewhere, usually a deep buried town water system, so it’s above freezing entering the house. If during its trip through the pipes it loses enough heat, it will freeze and form ice crystals. Unless those crystals attach to something, you get slush out the tap and the warmer water keeps feeding the system. If the crystals attach to the walls of th pipe they can start to build up and constrict the flow. if the flow from the town water supply is not warm enough to stop or reverse the buildup, it will eventually freeze solid.
But that new water flowing in the pipes has to lose, say, 84cal/g to get from liquid at nominal 4°C to a 0°C ice crystal. I would say water at 10C with a passable flow would need to be in some mighty cold environment to freeze before it hit the drain…
I wasn’t counting the heat of fusion, since I thought the whole point was that we were trying not to freeze. The water at the inner surface of the pipe is going to freeze eventually unless there’s enough flow to ensure that doesn’t happen. At a high enough flow, sure. But the OP asked about a drip.
I suppose that even a drip might be enough if you accept that the pipe will constrict somewhat, and that the layer of ice will act as insulation. Maybe it constricts down to a few millimeters inner diameter, which would increase the flow relative to the inner surface area. At a certain point, that might be enough to prevent it from freezing all the way through.
All in all though, there’s a lot of speculation in this thread and not a lot of hard data. Seems like someone must have done these experiments to see what the actual process is.
Throwing this out as I live in an area far enough north where pipes have been known to freeze and burst.
Common knowledge is if your pipes have the potential to freeze is to let them dribble overnight while you are sleeping. The water in the pipe in the ground below the frost line is +40 degrees. The part of the pipe that has the potential (on most houses) is only several feet long where it comes out of the ground, runs along a basement/crawl space wall, up an uninsulated outside wall, etc… The water may start to freeze in the pipe but if you went to bed around 10:00 pm and flushed the toilet and brushed your teeth several gallons of “warm “ water will have flowed through the line and “warmed” the water and pipes. All night long a little water flow prevents it from freezing shut. Then at 6:00am you wake up, use the toilet, shower, etc… and a lot of “warm” water flows through the pipe and melts any ice that may have been building up in the pipe. Obviously if you had cold weather for several days and you were not home, sooner or later your pipe would slowly freeze shut depending on how much water you had dribbling out.
A friend of the family had an older house with the water line buried too shallow where it came into the house so he had a lawn sprinkler timer in his basement that would turn on for a short time every few hours to flush the warmer water through the incoming water pipe.
This seems pretty plausible. A slow drip doesn’t seem like enough to keep the water in the pipe from freezing, and it mostly isn’t. It’s just enough to keep a narrow channel from freezing for long enough for greater water use to thaw the rest of it.
I also wonder how much tiny ice crystals that do form get pulled along by the slow drip into a warmer part of the house, which is a way that heat of solidification would contribute to keeping the rest of the pipe warm enough to be liquid.
As a black box, though. They know that doing this works, but we’re all speculating on exactly why it works.
As iamthewalrus_3 noted, the question isn’t whether it works at all, but why it works. So far we seem to have three basic categories of explanation:
The flowing water introduces heat from the relatively warm municipal supply
Flowing water doesn’t freeze
The open tap acts as pressure relief
The actual answer may be a combination of these. I posit that the last effect is dominant; the first may contribute under some conditions but is a relatively small factor; and the second is basically nonsense. Nothing beats an actual experiment, though.
I think the point is that for water to freeze it must go to 32 degrees and also lose enough energy to state change. Flowing water from a well or municipal source starts at a higher temperature and (usually) travels through the pipes before it has a chance to lose enough energy. Practical experience shows that the colder it gets, the higher the flow needed to prevent freezing which makes sense.
Still water has plenty of time to lose enough energy to state change.